OU Data Scientist Releases Free Tool for Studying Human-AI Interaction Research

The University of Oklahoma (OU) has released a free, open-source software tool designed to help researchers systematically study interactions between humans and artificial intelligence systems. Developed by Kat Gebauer, a data scientist at OU’s School of Library and Information Studies, the software aims to address gaps in current methods for analyzing how people engage with AI—whether in decision-making, creativity, or ethical dilemmas.

The University of Oklahoma

Researchers in human-AI interaction (HAI) studies often rely on ad-hoc methods to collect and analyze data, leading to inconsistencies in how human behavior is measured alongside AI outputs. Gebauer’s tool, unveiled on May 18, 2026, provides a standardized framework for capturing both qualitative and quantitative aspects of these interactions, including user trust, bias detection, and cognitive load.

The software integrates modular components for data collection (e.g., eye-tracking, survey responses, and task performance logs) with visualization tools to map interaction patterns. It is designed to be accessible to non-technical researchers, lowering the barrier for interdisciplinary teams in psychology, ethics, and computer science.

OU’s Health Campus, ranked in the Top 100 for NIH funding in health-related research, will initially pilot the tool in studies examining AI-assisted medical diagnostics. Gebauer emphasized that the software’s open-source nature allows researchers to adapt it for domains beyond healthcare, such as education or customer service AI.

“Current tools either focus on AI performance alone or treat human interactions as a black box,” Gebauer said in a statement. “This software bridges that gap by treating the human-AI dyad as a single system to study.”

The release aligns with growing demand for rigorous HAI research as AI systems become ubiquitous in professional and personal contexts. While proprietary platforms like Microsoft’s Copilot or Google’s Vertex AI offer interaction analytics, they typically require significant infrastructure investment and lack transparency in methodology. Gebauer’s tool, by contrast, is freely available on GitHub under an open license, with documentation targeting both academic and industry researchers.

OU’s announcement follows recent advancements in HAI research, including the National Science Foundation’s 2025 funding push for “human-centered AI” projects. The university’s Carnegie R1 designation and its School of Library and Information Studies—known for its work in information science—position it as a key contributor to this emerging field.

Next steps for the project include developing a community-driven update cycle, with input from researchers worldwide. Gebauer’s team has also begun collaborating with the university’s Office of Research to explore partnerships with tech companies for real-world validation.

For researchers seeking to adopt the tool, OU provides a dedicated support portal with tutorials, sample datasets, and a forum for troubleshooting. The software’s modular design allows users to customize modules for specific use cases, such as analyzing bias in generative AI outputs or measuring user fatigue in prolonged AI interactions.

As AI systems increasingly shape societal outcomes, tools like Gebauer’s will play a critical role in ensuring that human-AI interactions are studied with the same rigor as the AI systems themselves. The open-source approach also reflects a broader trend in research infrastructure, where transparency and accessibility are prioritized over proprietary control.